Effect of Ligand Modification on the Mechanism of Electrocatalytic

Feb 28, 2018 - Without computationally generated kinetic data, we must rely on electrochemical data to support this claim. Reduction potentials are sh...
0 downloads 10 Views 710KB Size
Subscriber access provided by UNIV OF SCIENCES PHILADELPHIA

Article

Effect of Ligand Modification on the Mechanism of Electrocatalytic Hydrogen Production by Ni(pyridinethiolate)3- Derivatives Carolyn Noelle Virca, Joshua R. Lohmolder, Joshua B. Tsang, Matthew M. Davis, and Theresa M. McCormick J. Phys. Chem. A, Just Accepted Manuscript • DOI: 10.1021/acs.jpca.7b11912 • Publication Date (Web): 28 Feb 2018 Downloaded from http://pubs.acs.org on March 2, 2018

Just Accepted “Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted online prior to technical editing, formatting for publication and author proofing. The American Chemical Society provides “Just Accepted” as a service to the research community to expedite the dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully peer reviewed, but should not be considered the official version of record. They are citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or consequences arising from the use of information contained in these “Just Accepted” manuscripts.

The Journal of Physical Chemistry A is published by the American Chemical Society. 1155 Sixteenth Street N.W., Washington, DC 20036 Published by American Chemical Society. Copyright © American Chemical Society. However, no copyright claim is made to original U.S. Government works, or works produced by employees of any Commonwealth realm Crown government in the course of their duties.

Page 1 of 35 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

The Journal of Physical Chemistry

Effect of Ligand Modification on the Mechanism of Electrocatalytic Hydrogen Production by Ni(pyridinethiolate)3- Derivatives C. N. Virca, J. R. Lohmolder, J. B. Tsang, M. M. Davis, T. M. McCormick* Portland State University, 1825 SW Broadway, Portland, OR 97232 Density Functional Theory, catalysis, proton reduction, nickel, electrochemistry

Abstract

The effects of ligand modification on the catalytic mechanism of hydrogen production by Ni(PyS )3- derivatives, made with electron withdrawing and donating substitutions to the pyridinethiolate (PyS) ligands, are studied experimentally and computationally using DFT. Thermodynamic data, spin density maps, and frontier molecular orbital diagrams were generated for reaction intermediates. Comparison of computed values for E0 and pKa with experimental values supports the proposed mechanisms. The rate of electrochemical hydrogen production is correlated with the effect of ligand modification. Notably, the presence of an electron donating substituent favors an alternative mechanism for hydrogen production. Computationally it was determined that the electron donating substituent causes deviation from the original Chemical-Electrochemical-Chemical-

ACS Paragon Plus Environment

1

The Journal of Physical Chemistry 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Electrochemical (CECE) mechanism of Ni(PyS)3-

Page 2 of 35

to a CCEE mechanism, while the

CECE mechanism is maintained for all catalysts substituted with electron withdrawing groups. Introduction Computational design of molecular hydrogen evolution catalysts made from earth abundant materials is an active area of research with significance for driving the development of effective renewable-energy storage technologies.1–10 Rationally designed homogenous catalysts can be optimized to operate under ideal conditions, at neutral pH and with low overpotentials.11–13 Towards this goal, the use of computationally driven catalyst design aids to elucidate reaction mechanisms of catalysts by determining the pKa’s and reduction potentials of intermediates for each possible protonation and reduction step along the catalytic cycle. Knowledge of a catalytic mechanism opens the possibility of rational catalyst modification. Given defined experimental parameters such as pH, the energy change of each step can be calculated and used to generate reaction coordinate diagrams.8,14 Mechanistic insights are obtained through study of the molecular geometry, energy, and electronic structure of reaction intermediates.1,8,9,15–21 Nickel-pyridinethiolate (Ni(PyS)3-), and simple derivatives of this compound, have been shown to be effective hydrogen evolution catalysts, under both photocatalytic and electrocatalytic conditions, capable of operating at pH 12 and with relatively low overpotentials.22,23 Our previous computational work on Ni(PyS)3- supports the catalytic Chemical-Electrochemical-Chemical-Electrochemical (CECE)

mechanism of hydrogen

production put forth by Eisenberg and co-workers under the given experimental conditions,24 as calculated pKa values and reduction potentials correlate well to those determined experimentally. In this work, we build on our previous study by probing how pKa and E0 are affected by

ACS Paragon Plus Environment

2

Page 3 of 35 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

The Journal of Physical Chemistry

modifying the pyridine ligand moiety with electron donating and withdrawing substituents spanning a range of Hammett constants. It has been previously shown that a methyl-substituted catalyst results in increased photocatalytic hydrogen production rates, in this work we suggest an alternative mechanism for this derivative.22 The investigated substituents are as follows: 3(trifluoromethyl)pyridine-2-thiol (3-CF3PyS), 5-(trifluoromethyl)pyridine-2-thiol (5-CF3PyS),22 6-mercaptopyridine-3-carboxylic acid (6-S-3-COOHPyS), 2-mercaptopyridine-3-carboxylic acid (2-S-3-COOHPyS), 5-chloropyridine-2-thiol (5ClPyS),22 6-methylpyridine-2-thiol (6-CH3PyS)22 (structures in Figure 1 and Table 1). The catalytic cycles of these compounds were studied computationally and the complexes were subsequently synthesized and characterized experimentally. The previously supported CECE mechanism of Ni(PyS)3- serves as the starting point for analysis of the catalytic cycle. The most notable finding of our study is that for the most electron donating substituent (F), agreement between experimental and theoretical data suggests a unique CCEE mechanism. Initial protonation and reduction steps are studied by visualization of the frontier molecular orbitals, spin density maps, pKa values of various sites, E0 values, and thermodynamic reaction coordinate diagrams for reaction intermediates. Remaining protonation and reduction steps are discussed and the effect of pKa and E0 on experimentally determined turnover frequencies (TOFs) is proposed. We use this data to investigate the relationship between E0, pKa and σ (Hammett constant), used to quantify the effect of the electron donating/withdrawing ability of the ligand substituent.

ACS Paragon Plus Environment

3

The Journal of Physical Chemistry 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 4 of 35

Figure 1: Schematic of Ni(PyS)3- compound, R groups are given in Table 1 Table 1: Specified substitutions on Ni(PyS)3- schematic with associated Hammett constants55. R

R’

R’’

σ

(3-CF3PyS)

-H

-H

-CF3

0.43

(5-CF3PyS)

-H

-CF3

-H

0.43

(6-S-3-COOHPyS)

-H

-COOH

-H

0.37

(2-S-3-COOHPyS)

-H

-H

-COOH

0.37

(5-ClPyS)

-H

-Cl

-H

0.37

(6-CH3PyS)

-CH3

-H

-H

-0.07

Methods Computational Methods DFT calculations were

performed

using

Gaussian

09

following

well-established

procedures.9,13,25 Optimizations were run using the B3P8626–28/6-31+G(d) level of theory with the CPCM29,30 solvent model for water–based on previous studies.24 A comparison of the metric parameters around the nickel ion for each starting catalyst, relative to the unsubstituted Ni(PyS)3, can be seen in Table S1 (Figure S1). Normal mode frequency calculations were performed to confirm computed minima and to provide vibrational modes for thermochemical analysis.

ACS Paragon Plus Environment

4

Page 5 of 35 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

The Journal of Physical Chemistry

Reduction potentials were computed using isodesmic reactions with the Ni(I)(PyS)3H/ Ni(II)(PyS)3H- couple as the reference reaction (Eref = –1.62 V vs SCE) to generate a balanced redox reaction and calculate ∆Giso used in the equation below. Use of a reference reaction in this manner reduces systematic errors resulting from electron exchange functionals and choice of basis set.9 The pKa values were determined by balancing chemical equations with a solvated proton that has an energy value of -264 kcal/mol to account for the required dechelation.16 Values for E0 and pKa were calculated using ∆G values for given reductions and protonations using the following equations: ∆Grxn = [ΣEprod] – [ΣEreact] E0 vs SCE = (–∆Giso/nF) + Eref pKa = -∆G/[ln(10)RT] where F is the Faraday constant and n is the number of electrons (1 in this case). ∆Giso is the change in Gibbs free energy of the isodesmic reaction9 describing electron exchange between the reduced (Red) and oxidized catalyst (Ox) with the unsubstituted catalyst: Ni(PyS)3H- + Ox  Ni(PyS)3H + Red Eref is the value of the selected reference reaction; in this case the experimental Ni(PyS)3H/Ni(PyS)3H couple referenced against SCE (cyclic voltammogram of Ni(PyS)3- reference in Figure S2). Thus all values of E0 presented herein are referenced vs. SCE. Experimental Methods Compounds Ni(5-CF3PyS)3-, Ni(5-ClPyS)3-, and Ni(6-CH3PyS)3- were synthesized as previously described (Table 1).22 Compounds Ni(3-CF3PyS)3-, Ni(6-S-3-COOH)3-, and Ni(2-S-3COOHPyS)3- were synthesized using a similar procedure, where a solution of deprotonated ligand was prepared using sodium methoxide in dry/degassed methanol. Following color change,

ACS Paragon Plus Environment

5

The Journal of Physical Chemistry 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 6 of 35

indicative of deprotonation, an equivalent of [Et4N]Br was added to exchange the counter ion. Solvent was removed in vacuo and the remaining solid was dispersed into dry/degassed acetonitrile, to remove the residual salt by filtration. A solution of [Et4N]2NiCl4 in acetonitrile was added slowly over the course of 30 minutes and allowed to react for at least two hours, until green product was formed. The catalyst then precipitated out which allowed us to remove residual salts through filtration following reduction of the solution volume by about half. Reduction potentials were obtained from cyclic voltammograms of each complex in dry/degassed acetonitrile solutions with concentrations on the order of 10-3 M of catalyst, using a Ag/AgNO3 reference electrode. Overpotentials were determined by first calculating the standard ° reduction potential (E/ ) of 4-cyanoanilinium using the following equation31–33:  ° ° E/ = E / −  

2.303RT  pK ,, F

° Where E / is the standard potential for the solvated proton and dihydrogen couple in a 

given solvent. We use the value of -0.028 V vs Fc+/Fc in acetonitrile.34 The over potentials, η, is  the different of E  − E/ . Prior to addition of acid, a reduction peak was not observed within 

the solvent window for any of the derivatives. Following addition of acid, a reduction peak was observed (one equivalent of acid was added to all derivatives except Ni(6-S-3-COOHPyS)3- and Ni(2-S-3-COOHPyS)3-, which needed four equivalents to ensure protonation of the carboxylic acid groups). Subsequent addition of acid led to a catalytic wave in all cases. These traces can be seen in Figure S3-8. TOF of proton reduction was measured by plotting the ratio of icap/ip vs concentration of acid, until the ratio approached a plateau (icat = catalytic current, and ip = initial current). In the region where there is no longer a linear dependence on the concentration of acid, the ratio of icat/ip was used to calculate kobs (observed rate constant) and the TOF. The process is described in detail in

ACS Paragon Plus Environment

6

Page 7 of 35 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

The Journal of Physical Chemistry

the experimental section, using the approximate model for a pseudo-first order catalytic system.33,35–38 The shape of the catalytic curve indicates that pure kinetic conditions with substrate consumption are never reached. It is often challenging to reach pure kinetic conditions due to the occurrence of multiple reaction mechanisms, especially in cases where isomers exist.39 Extracted kobs are used for internal comparison between the selected catalysts in this study. This method has previously been used to estimate the rates (TOF’s) of systems with complex catalytic mechanisms.2,40–45 Electrocatalytic TOF’s determined using this method (Table 2) follow the same trends as the photocatalytic TOF previously determined by Eisenberg and coworkers.22 To determine the pKa’s of each derivative, solutions in 1:1 H2O/EtOH with catalyst concentration of the order of 10-4 M were fully deprotonated using aqueous NaOH and the UVvis absorbance of the deprotonated catalysts was determined. Samples were titrated with aqueous HCl and the pH and absorbance spectra were monitored throughout, until complete conversion to the protonated catalyst as noted by no further change in the absorbance spectra. Absorbance spectra of the fully protonated and fully deprotonated catalysts can be seen in Figure S9 as well as the titration spectra and accompanying plots of the absorbance ratio vs pH in Figure S3-S8. The pKa was chosen to be the minimum value of the second derivative of the curve of best fit of the ratio of λmax of the protonated and deprotonated catalysts plotted against pH, which corresponds to its inflection point. Not all of the plots of UV-vis absorbance spectra show a clear isosbestic point in moving from the protonated to the deprotonated species in solution. We attribute this to the existence of three existing isomers of each catalyst in solution. Current work is ongoing to further investigate the effects of having multiple isomers of nickel-based proton reduction catalysts. By using the experimental properties of each derivative we were able to

ACS Paragon Plus Environment

7

The Journal of Physical Chemistry 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 8 of 35

Table 2: Computationally determined pKaŧ values and reduction potentials* (V vs SCE) with experimental pKa’s, reduction potentials, TOFs and overpotentials (ηexp). N

pKa S

pKa

TOF s- ηexp

Ligand

(calc)

(calc)

pKa (exp)

E0(calc)

E0(exp)*

(3-CF3PyS)

6.3

-7.7

8.3

-1.34

-1.26

42

325 mV

(5-CF3PyS)

6.6

1.5

7.4

-1.30

-1.29

53

354 mV

1

194 mV

(6-S-3COOHPyS) 9.9

-0.2

10.3

-1.36

-1.13

164

(2-S-3COOHPyS) 7.0

4.5

8.6

-1.37

-1.25

164

7.8

4.0

7.6

-1.38

-1.31

314

375 mV

7.6

10.7

-1.55

-1.09

332

155 mV

(5-ClPyS)

(6-CH3PyS) 16.3

315 mV

ŧ

pKa values calculated from the triplet state. *Electrochemical data in Figures S3-S8. undertake a more focused computational investigation into the mechanism of hydrogen production by these compounds. Results and Discussion First chemical step The first step of the catalytic cycle of hydrogen production by Ni(PyS)3- was previously determined to be protonation.24 For each derivative reduction prior to protonation would require large potentials (Table S2). Under our experimental conditions, reduction was observed for all catalysts using an applied potential of -1.58 V vs SCE or less only after a proton source was added (Figures S10), thus protonation was considered to be the first step in the catalytic cycle of all derivatives. The catalyst could be protonated at a pyridyl N, or on a S atom. In our previous work exploring the catalytic cycle of Ni(PyS)3-, we supported the experimentally determined site of protonation by comparing ∆G for protonation at various possible sites and selecting the most

ACS Paragon Plus Environment

8

Page 9 of 35 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

The Journal of Physical Chemistry

thermodynamically favored option. For Ni(PyS)3- this was found to be on the dechelated pyridyl N. Furthermore, protonation at the pyridyl N is supported by a previously reported single crystal X-ray structure of a similar compound, [Ni(HpyS)4](NO3)2·EtOH, which has all four ligands protonated at the pyridyl N and dechelated.22 The computed pKa values for nitrogen and sulfurbased protonation can be seen in Table 2. For all six derivatives, protonation at the pyridyl N is thermodynamically favored relative to a S atom. Our computed values pKa generally agree with experimental data, with the exception of Ni(6-CH3PyS)3-. For Ni(6-CH3PyS)3- neither the computed pKa for protonation of N nor S matches the experimentally obtained value. This inconsistency suggests a deviation from the CECE mechanism. We have found that Ni(6-CH3PyS)3- follows a unique catalytic mechanism due to the electron donating effect of the methyl substitution on the pyridinethiolate ligands. Although the Hammett constant of a methyl substituent does not differ greatly relative to the Hammett constant of a H substituent, it appears that even a minor increase in electron density is enough to pass a threshold beyond which a new mechanism becomes accessible. Interestingly, this catalyst has the highest electrochemical TOF of 332 s-1 (plots of icat/ip used to generate the TOF’s can be seen in Figure S3-S8). This follows the previously reported photocatalytic activity of this catalyst that was found to have the highest TOF out of a series of nickel pyridinethiolate complexes.22 We investigated the possible protonation sites for Ni(6-CH3PyS)3-. The cycle begins with Ni(6CH3PyS)3- in a triplet-spin-state, denoted {t}. The triplet-spin-state is selected over the singletspin-state, {s}, as it is the lower energy conformation and has the appropriate octahedral geometry (Figure 2). As with the other derivatives, protonation of Ni(6-CH3PyS)3-{t} could be the first step, occurring at either a S or N of the pyridinethiolate ligand. The calculated pKa’s are

ACS Paragon Plus Environment

9

The Journal of Physical Chemistry 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 10 of 35

7.6 for S protonation and 16.3 for N protonation, indicating that with experimental conditions at pH 7 both of these sites can be protonated (Table S2). The intermediate protonated at a N is more thermodynamically favored with a ∆G at pH 7 of -12.7 kcal/mol compared to the ∆G for protonation at a ligand S atom of only -1 kcal/mol. Following initial protonation, Ni(6-CH3PyS)3HN{t} can either undergo a second protonation at a S, N or Ni, or be reduced to either Ni(6-CH3PyS)3H-{d} or Ni(6-CH3PyS)3H-{q}. Protonation at S and Ni have respective pKa’s of 1.0 and -2.6, as such, neither species will be significantly present at pH 7. However, dechelation and protonation of a second ligand N corresponds to a pKa of 10.44 (∆G = -4.7 kcal/mol at pH 7) (Figure 2). The experimentally measured pKa for Ni(6CH3PyS)3- is 10.7, suggesting that proceeding through a doubly protonated intermediate could be a viable pathway. The pKa corresponding to the first protonation cannot be observed using this method as the highest pH's of the solutions used during experiment are between 13-14. Given the computed initial pKa of this compound, it will already be protonated in solution before the measurement begins. Due to the method by which we determined pKa’s experimentally, without the application of an externally applied potential, the measured pKa of 10.7 must necessarily correspond to a step that occurs prior to reduction. Thus, we propose that two sequential protonation events start the catalytic cycle, given experimentally determined pKa. A CCEE mechanism was investigated for the other derivatives of Ni(PyS)3- as well, but due to the results of those computations can be found in Table S2.

ACS Paragon Plus Environment

10

Page 11 of 35

50

40

e30 ΔG (kcal/mol)

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

The Journal of Physical Chemistry

20

H+ shift

10

0

H+

e-

-10

H+ -20

Figure 2: Reaction coordinate diagram of the proposed catalytic cycle of hydrogen production by

Ni(6-CH3PyS)3- showing CCEE mechanism with proton shift from a ligand nitrogen to the Ni center (at pH 7). After undergoing two protonations, Ni(6-CH3PyS)3HNHN+{t} can be reduced to Ni(6CH3PyS)3HNHN{d} or Ni(6-CH3PyS)3HNHN{q}. The double-state has electrons localized on the metal, while the quartet-state would have electrons localized on a ligand. E0calc for reduction to Ni(6-CH3PyS)3HNHN{d} is -0.89 V, while E0 to Ni(6-CH3PyS)3HNHN{q} is -2.16 V (Table S1). Experimentally, the measured reduction potential is -1.09 V vs SCE. Thus, reduction to the doublet-state, with the electrons on the metal center, is preferred. From Ni(6-CH3PyS)3HNHN{d}, the necessary second reduction to yield Ni(6-CH3PyS)3HNHN-{s} or Ni(6-CH3PyS)3HNHN-{t} requires -2.80 V and -2.58 V, respectively, making both reductions too energetically demanding to occur under the electrochemical hydrogen production conditions used herein (Figure S11). Thus suggests that hydrogen production is occurring through reduction a different intermediate.

ACS Paragon Plus Environment

11

The Journal of Physical Chemistry 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 12 of 35

An intermediate with a proton on the Ni center, Ni(6-CH3PyS)3HNHNi{d} has a calculated reduction potential of -1.18 V vs SCE (Figure 2), which is in agreement with the experimentally determined potential of -1.09 V vs SCE. Although direct protonation of the metal center is not favored (pKa of -2.6), proton shifts have been previously reported in proton reduction catalytic cycles; from pendant amines to Ni centers in NiP2N2 type catalysts.46 We propose that an internal proton shift occurs from a protonated ligand N to the metal center for Ni(6-CH3PyS)3-. This shift requires 21 kcal/mol, and leads to the formation of Ni(6-CH3PyS)3HNHNi{d}. The transition state corresponding to this movement of the proton from the ligand to the metal center was optimized (Figure 3) and the reaction coordinate diagram of this can be seen in Figure S12. In order to circumvent the large energetic requirement, we propose the proton shift and reduction occur simultaneously. Without computationally generated kinetic data we must rely on electrochemical data to support this claim. Reduction potentials are shifted negatively by approximately 10 mV in deuterated solvent vs protonated solvent, suggesting PCET is occurring (Figure S13). This mechanism accounts for both the experimentally measured pKa and reduction potential and helps to explain why Ni(6-CH3PyS)3- ultimately has the lowest overpotential and highest TOF of the derivatives that were studied. Ni(6-CH3PyS)3- can be readily protonated twice at pH 7 and subsequently be reduced at lower potentials than the other derivatives. The CCEE mechanism of Ni(6-CH3PyS)3- contains the smallest energy fluctuations of all of the hydrogen production pathways considered (Figure S14).

ACS Paragon Plus Environment

12

Page 13 of 35 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

The Journal of Physical Chemistry

Figure 3: Optimized geometry of NiF3HNHN/Ni{d} transition state. One final point of consideration for Ni(6-CH3PyS)3- is the position of the electron-donating group on the pyridyl ring. The electron-donating group is in the ortho position relative to the N, while electron withdrawing substitutions we studied are in the meta and para positions for compounds Ni(3-CF3PyS)3--Ni(5-ClPyS)3-; these derivatives were chosen due to commercial availability of the ligands. In order to ensure that steric effects are not responsible for the unique mechanism of this catalyst, analogous computations were performed with the donating group at the meta and para position relative to the pyridyl N, to minimize steric hindrance. The corresponding pKa’s for the protonated meta and para substituted derivatives are 14.6 and 15.1, respectively compared to 16.3 for Ni(6-CH3PyS)3-. Though there is an effect on pKa as a result of substitution position, all three of these compounds are still significantly more basic than compounds Ni(3-CF3PyS)3--Ni(5-ClPyS)3-, whose pKa’s range from 6.3 to 7.8. Subsequent protonation required for the CCEE mechanism is possible again for both isomers at pH 7 (pKa para = 9.3 and pKa meta = 7.1). The remaining two electrical steps occur for the meta isomer at -0.95 V vs SCE and -1.10 V vs SCE, and at -1.04 V vs SCE and -1.10 V vs SCE for the para isomer (Table S1). These values are similar to those calculated for Ni(6-CH3PyS)3-ortho of -0.89 V vs

ACS Paragon Plus Environment

13

The Journal of Physical Chemistry 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 14 of 35

SCE and -1.18 V vs SCE. Finally, the energetic requirement for the proton shift from the pyridyl N to the Ni center for both of these isomers are within 1 kcal/mol of the 21 kcal/mol requirement for Ni(6-CH3PyS)3-. Thus, regardless of the position of the electron donating substituent, its impact on the catalytic mechanism is maintained. However, for electron-withdrawing –COOH substitutions the position has a significant effect on the pKa of the third intermediate. For compounds Ni(6-S-3-COOH)3- and Ni(2-S-3-COOH)3-, which contain –COOH in para and ortho positions, respectively relative to the sulfur, Ni(6-S-3COOH)3- has a pKa of -1.80 and Ni(2-S-3-COOH)3- has a pKa of 9.23 (Table 3). We suspect the difference in acidity for these two compounds is a result of the proximity of the ortho –COOH substituent to the metal center which can stabilize the protonated state by acting as an intramolecular proton shuttle. Work studying the effects of pendant amines on Ni containing catalysts has shown that pendant ligands, when in the vicinity of the metal center, can have stabilizing effects on critical reaction intermediates such as hydrogen and hydride containing complexes.14,47–49 This leads to significant increases in the pKa for stabilized intermediates. An

Table 3: Second pKa (with location of protonation) and second reduction potential (V vs SCE)

Ligand

pKa (calc) (2)

E0(calc) (2)

(3-CF3PyS)

2.35 (Ni)

-0.96

(5-CF3PyS)

2.38 (Ni)

-1.04

(6-S-3-COOHPyS)

-1.80 (Ni)

-0.88

(2-S-3-COOHPyS)

9.23 (Ni)

-0.94

(5-ClPyS)

8.86 (Ni)

-1.38

(6-CH3PyS)

10.44 (N)

-1.18

ACS Paragon Plus Environment

14

Page 15 of 35 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

The Journal of Physical Chemistry

increase in basicity of 8 pKa units has been reported for isomers of a hydrogenase models that contain amine/metal center interactions.32 This same effect of position on pKa is not observed when comparing the –CF3 derivatives, Ni(3-CF3PyS)3- and Ni(5-CF3PyS)3- that have very similar pKa values of 2.35 and 2.38 respectively. Ligand involvement is not restricted to protonation events. Non-innocent ligands can also influence reduction. In the CECE mechanism, the compound is reduced after the first protonation. Our intuition suggests reduction of compounds with decreased electron density may be ligand centered rather than metal centered. The first reduction in the catalytic cycle brings the compound back to an overall -1 charge and could result in a doublet-spin-state if reduction occurs at the metal center, or a quartet-spin-state, for which electron density is distributed on the protonated ligand. The parent compound, Ni(PyS)3-, was determined to undergo reduction to a low spin, doublet state, where the incoming electron is localized on a d-orbital at the nickel center (Figure S15). The LUMO (designated as 100 β-orbital) is the logical recipient for this first electron (Figure 4). From visual inspection of this unoccupied frontier molecular orbital, it is

101-α

101-β

100-β LUMO

101-α HOMO 100-α

100-β

99-α

99-β

99-α

100-α

Figure 4: MO diagrams from Ni(PyS)3H for orbitals 99, 100 and 101, showing that the LUMO, 100-β, is localized at the nickel center, which would lead to a low spin configuration of

Ni(PyS)3H- following reduction.

ACS Paragon Plus Environment

15

The Journal of Physical Chemistry 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 16 of 35

apparent that a portion of the orbital is indeed the result of contributions from d-orbitals of the nickel center. Furthermore, calculated values of E0 are systematically higher for reduction to the quartet-spin-state, resulting from reduction of the ligand, than for the metal-centered reduction to the doublet-state. The reduction step was further investigated by studying the occupied orbitals of the reduced compound. Frontier molecular orbitals of Ni(PyS)3H- for the low-spin state show the location of electron density centered around the nickel (Figure S16). Given that the low-spin complex is systematically lower in energy than the high-spin analogue for the parent and all derivatives studied, we conclude that the resulting spin-state after the first electrochemical step is a doublet, indicating that non-innocent ligands are not involved even for electron poor ligands. MO diagrams of SOMO 100 for each reduced derivative indicate that a portion of the electron density for the low-spin compounds is indeed found on the metal center as it is for the parent compound (Figure S16). A second protonation and reduction must occur to produce hydrogen. For all compounds, except Ni(5-ClPyS)3- and Ni(6-CH3PyS)3-, the first calculated reduction potential corresponds to the experimentally measured E0 value. Reduction of Ni(5-ClPyS)3- was calculated at -1.38 V for both the first and second reduction, making it impossible to distinguish which event is being measured experimentally. In our initial computational investigation of Ni(PyS)3- catalyzed proton-reduction, the second reduction is attributed to formation of a Ni-hydride through PCET.24 The CECE pathway was modeled with placement of the second proton on the Ni center. The pKa’s for these species can be seen in Table 3. For Ni(3-CF3PyS)3-, Ni(5-CF3PyS)3- and Ni(6-S3-COOH)3- highly acidic compounds result, indicating that for these derivatives this chemical

ACS Paragon Plus Environment

16

Page 17 of 35 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

The Journal of Physical Chemistry

step may be occurring via PCET to yield a hydride intermediate, as was the case for Ni(PyS)3-. It has been noted previously that when high-energy intermediates result from proton transfer steps, the energetic barrier can be bypassed by concerted addition of a proton and an electron simultaneously.14 Computational investigation of PCET steps were not considered in this study. Rather we look at the second protonation and reduction as two separate steps to distinguish the thermodynamic requirements of each. Thus, the final step is a reduction, which brings all compounds back to an overall -1 charge and results in formation of an intermediate that can exist as either a triplet or a singlet-spin-state. A singlet-state would require the electron to be localized on the metal center, while a triplet-spinstate could be a result of electron density located on either the H atom or a PyS ligand. Spin density maps of the unpaired electron density supports formation of a hydride (Figure S17). The triplet-states for most of the compounds are lower in energy than their singlet-state analogues (Table S2). The triplet-state of Ni(5-CF3PyS)3- is slightly higher in energy than the singlet-state, however, the singlet-state intermediate has complete dechelation of the protonated ligand. Furthermore, the proton and hydride are not spatially oriented for elimination of hydrogen in the singlet-state-intermediate, thus we suggest that the triplet-state-intermediate is still part of the reaction pathway.

ACS Paragon Plus Environment

17

The Journal of Physical Chemistry

-1.25 -1.35 -1.45 E0

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 18 of 35

-1.55 -1.65 -1.75 -0.1

0

0.1

0.2

0.3

0.4

0.5

σ (R')

Figure 5: Relationship between E0 from computed data and Hammett constants. Additional CCEE mechanism considerations For Ni(6-CH3PyS)3- it is possible that protonation occurs at the N atom trans to the protonated N. The previously discussed results in the section describing Ni(6-CH3PyS)3- highlight results from protonation of the N atom located cis to the already protonated N. The pKa values for these two protonation sites differ significantly, due to each N atom being in a unique position relative to the other heteroatoms. Optimized structures of both of these doubly protonated compound can be seen in Figure S18. The pKa value for the trans protonated compound is 26.9, which likely results from stabilizing hydrogen-bonding interactions with the nearby sulfur atom. Reductions following this trans protonation event occur at reasonable potentials (Table S2). However, because the pKa value is not directly supported with experimental data, it was not highlighted alongside the previously described CCEE mechanism, though it is likely accessible given the experimental conditions used. It is also important to mention our investigation of a CCEE mechanism for the remaining derivatives of Ni(PyS)3-. The results of these computations can be seen in both Table S2 and

ACS Paragon Plus Environment

18

Page 19 of 35 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

The Journal of Physical Chemistry

Figure S18, which contain the energies, optimized structures and pKa values of both the trans and cis protonations. As mentioned previously, several isomers exist for each compound and both protonation sites are unique, with different pKa values. For example, for Ni(3-CF3PyS)3- the trans protonation site has a pKa of 4.6 while the cis position has a pKa of 6.0. Computational results alone suggest that in some instances that CCEE mechanisms may be accessible, but due to a lack of experimental support for a CCEE mechanism, this pathway was not considered further. A more thorough computational investigation is required to provide support for this possible mechanism, along with new experimental methods. Investigation of a CCEE mechanism undergone by electron poor derivatives of Ni(PyS)3- is currently being undertaken to better understand the role that multiple isomers, with variable pKa values, may have on influencing the hydrogen production mechanism of these compounds. The relationship between initial pKa, initial E0 and Hammett constant is consistent with the literature.12 As substituents become more electron withdrawing, E0 decreases in magnitude, a desirable result to decrease required overpotentials (Figure 5). The reverse relationship is observed between pKa values and Hammett constant.

As electron-withdrawing capability

increases, pKa decreases, limiting the range of pH’s at which the protonated species can exist and begin the catalytic cycle. Consequently, pKa and E0 are inversely affected by substitution. This inverse relationship makes it difficult to optimize both the pKa and reduction potential using simple ligand substitution. However, as shown in the discussion of Ni(6-CH3PyS)3-, changes in the mechanism that result from ligand modification can supersede this relationship. Furthermore, this is observed in the influence of ligand modification on the experimentally measured electrochemical TOF.

ACS Paragon Plus Environment

19

The Journal of Physical Chemistry 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Catalytic

efficiency

was

quantified

through

TOFs,

Page 20 of 35

which

were

determined

electrochemically.33,50–53 It appears that the most electron withdrawing substitutions negatively impact the rate of hydrogen production, as seen in Table 2 where compounds Ni(3-CF3PyS)3and Ni(5-CF3PyS)3- have lowest experimentally measured TOF’s. The two carboxylic acid containing catalysts, Ni(6-S-3-COOH)3- and Ni(2-S-3-COOH)3-, show a mild increase in TOF relative to Ni(3-CF3PyS)3- and Ni(5-CF3PyS)3-. Although Ni(5-ClPyS)3- is substituted with Cl groups that have the same Hammett-constant as the –COOH groups, it has a TOF that is nearly double, 314 s-1 for Ni(5-ClPyS)3- compared to 164 s-1 for Ni(6-S-3-COOH)3- and Ni(2-S-3COOH)3-. The TOF of Ni(5-ClPyS)3- is quite similar to that of Ni(6-CH3PyS)3- (332 s-1), which has the most electron rich ligands. The effect of these substituents on the rate of hydrogen production warrants further investigation. Conclusions Through the joint use of computational and experimental means, six derivatives of Ni(PyS)3were studied to evaluate the effect of ligand modification on properties of the catalysts and their hydrogen production mechanisms. Compounds containing electron poor ligands maintain the CECE mechanism followed by the parent compound. The TOF’s of these derivatives are limited by the initial metal-centered reduction event, which occurs at a greater potential than the second reduction. E0 and pKa for the first two steps of these catalytic cycles show linear correlation with Hammett-constants and the relationship between these two parameters is inversely favorable. The electron rich ligand has the most substantial impact on the mechanism of hydrogen production, causing deviation from a CECE mechanism in favor of a CCEE mechanism, which involves a proton shift from a pyridyl N to the Ni center following the second protonation. Materials

ACS Paragon Plus Environment

20

Page 21 of 35 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

The Journal of Physical Chemistry

6-Mercaptopyridine-3-carboxylic (Trifluoromethyl)pyridine-2-thiol,

acid,

2-Mercaptopyridine-3-carboxylic

5-(Trifluoromethyl)pyridine-2-thiol,

acid,

3-

2-Mercapto-6-

methylpyridine, 5-chloropyridine-2-thiol, nickel (II) chloride, and sodium metal were purchased from Sigma Aldrich and used as received. Ferrocene were purchased from Acros Organics and used without purification. Lithium chloride was purchased from Fluka and used as received. Solvents were used without further purification. Mass spectroscopy was performed on a Thermo Electron LTQ-Orbitrap Discovery high-resolution mass spectrometer with electrospray ionization. General Synthetic Procedure All syntheses were performed using dry/degassed solvents under N2 atmosphere and complexes were stored under N2 atmosphere. The same procedure was used to synthesize all six compounds22 with the exception of an additional equivalent of sodium used in the synthesis of Ni(6-S-3-COOHPyS)3- and Ni(2-S-3-COOHPyS)3- to ensure deprotonation of both the carboxylic acid substituent and the thiol substituent. Ni(3-CF3Py-2-S)3Sodium metal (.07 mmols) was dissolved into 1 mL of dry/degassed methanol to yield the necessary solution of sodium methoxide. Into this solution 0.26 mmols of 3(trifluoromethyl)pyridine-2-thiol (ligand A) was added and allowed to dissolve resulting in a clear yellow solution. Then 0.9 mmols of [Et4N]Br was then added and allowed to react for one hour before solvent was removed by rotary evaporation and the remaining solid was dissolved into 0.8 mL of acetonitrile. Residual salt was removed through filtration before the solution was placed back under N2. A solution of 0.26 mmols of [Et4N]2NiCl4 in 0.8 mL of acetonitrile was slowly added via syringe over the course of 30 minutes. If the nickel precursor is added too

ACS Paragon Plus Environment

21

The Journal of Physical Chemistry 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 22 of 35

quickly, a brown solution results instead of the desired green solution. During the addition, if slight brown precipitate was observed, the addition could be slowed down and the reaction proceeded as intended. The final mixture was allowed to react for two hours, or until complete color change to blue/green occurred. The volume of solution was reduced by half prior to filtering to remove salt byproducts and was stored under N2 at -10°C to yield a solid product. FTMS–p ESI Calcd. for C18H9N3S3F9Ni, 591.9168; found, 591.8486. Ni(5-CF3Py2-S)3Sodium metal (3.4 mmols) was combined with 5 mL of dry/degassed methanol to make sodium methoxide. 5-(trifluoromethyl)pyridine-2-thiol (ligand B) (3.4 mmols) was added to this solution and yielded a clear yellow solution once dissolved, after which 3.4 mmol of [Et4N]Br was added and allowed to react for one hour. Following this, solvent was removed in vacuo and the resulting solid was dissolved into 10 mL of dry/degassed acetonitrile. Salts were filtered out and the remaining solution was placed back under N2 before a 10 mL solution of 0.85 mmol [Et4N]2NiCl4 in 10 mL of acetonitrile was slowly added over 30 minutes. Following the addition, complete color change to blue/green occurred and the volume of the solution was reduced by half before being stored at -10°C under N2 to yield a solid product. FTMS – p ESI Calcd. for C18H9N3S3F9Ni, 591.9168; found, 591.8523. Ni(3-COOH3Py-6-S)3Sodium methoxide was made by dissolving 2.88 mmols of sodium into 2.2 mL of dry/degassed methanol. Ligand C (6-mercaptopyridine-3-carboxylic acid) (1.44 mmols) was then added and allowed to react over the course of an hour, ultimately yielding a clear yellow solution into which 2.88 mmols of [Et4N]Br was added. This solution was allowed to stir for an hour before the solvent was removed in vacuo and the resulting solid was dissolved into 4.3 mL of

ACS Paragon Plus Environment

22

Page 23 of 35 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

The Journal of Physical Chemistry

dry/degassed acetonitrile. Salts were removed through filtration and the solution was placed under an atmosphere of N2. A solution of 0.36 mmols of [Et4N]2NiCl4 in 4.3 mL of acetonitrile was made and slowly added over the course of 60 minutes. The length of time of the addition was increased to an hour to avoid solubility issues resulting from the formation of the carboxylate. Once the reaction mixture appeared blue/green, the volume was reduced by half and the solution was placed under N2 at -10°C and the solid product was collected. FTMS – p ESI Calcd. for C18H10N3S3O4Na2Ni, 531.8983; found, 531.9774. Ni(3-COOH3Py-2-S)3A 2.2 mL solution of sodium ethoxide was made by combining 2.88 mmols of sodium with 2.2 mL of dry/degassed methanol. Ligand D (2-mercaptopyridine-3-carboxylic acid, 1.44 mmols) was added and allowed to stir for an hour to ensure deprotonation, at which point 2.88 mmols of [Et4N]Br was added and allowed to react for an hour. Solvent was then removed in vacuo and the resulting solid was dissolved into 4.3 mL of dry/degassed acetonitrile with residual salts removed by filtration. This solution was placed under N2 and over the course of 60 minutes a solution of 0.36 mmols of [Et4N]2NiCl4 in 4.3 mL of acetonitrile was added to yield a blue/green product. The final volume of solution was reduced to half in vacuo and stored under N2 at -10°C to yield a solid product. FTMS – p ESI Calcd. for C18H10N3S3O4Na2Ni, 531.8983; found, 531.9769. Ni(5-Cl3Py-2-S)3Sodium methoxide was made by dissolving 1.44 mmols of sodium into 2.2 mL of dry/degassed methanol and used to dissolve 1.44 mmols 5-chloropyridine-2-thiol (ligand E). Following addition of ligand E the solution immediately turned clear yellow and 1.44 mmols of [Et4N]Br was added and allowed to react for an hour. After an hour, the solvent was removed

ACS Paragon Plus Environment

23

The Journal of Physical Chemistry 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 24 of 35

and the resulting solid was dissolved into 4.3 mL of dry/degassed acetonitrile. Residual salts were filtered off and the solution was placed back under N2. Into this solution a 4.3 mL solution of [Et4N]2NiCl4 in acetonitrile was slowly added over the course of 30 minutes to yield a bright green solution of mixture, the volume of which was reduced by half before storing at -10°C under N2 to afford the solid product. FTMS – p ESI Calcd. for C15H9N3S3Cl3Ni, 491.8348; found 491.7782. Ni(6-H3Py-2-S)3Sodium metal (3.4 mmols) was dissolved into 5 mL of dry/degassed methanol to form a solution of sodium methoxide. Into this solution 3.4 mmols of 6-methylpyridine-2-thiol (ligand F) was dissolved and combined with 3.4 mmols of [Et4N]Br before being allowed to react for an hour. After an hour, the solvent was removed and the resulting solid was dissolved into 10 mL of dry/degassed acetonitrile. Salts were removed using filtration and the solution was placed back under N2. Over the course of 30 minutes a 10 mL solution of [Et4N]2NiCl4 in acetonitrile was added to yield a light green product. The volume was reduced by half and the final solution was placed under N2 at -10°C. Ni(6-CH3PyS)3- was isolated as a very light green solid. Calcd. for C6NSH6 (ligand F); 124.0221 found 124.0092. Elemental analysis for C, H and N calculated with one water molecule to be C; 53.863 measured 53.232, H; 7.128 measured 7.418, and N; 9.664 measured 9.233. Determination of E0 Cyclic voltammetry measurements were performed with a Biologic SP 200 potentiostat using a one-compartment cell with a glassy carbon working electrode, a Pt-wire counter electrode and a Ag+/AgNO3 acetonitrile reference electrode. A 10-3 M solution of catalyst made in dry/degassed acetonitrile was titrated with up to 30 equivalents of a 1.0 M solution of 4-cyanoanilinium in

ACS Paragon Plus Environment

24

Page 25 of 35 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

The Journal of Physical Chemistry

dry/degassed acetonitrile, with N2 flowing in the headspace of the container. The potential was first swept from 0.0 V vs Ag+/AgNO3 to -1.2 V vs Ag+/AgNO3, then back to 0 V vs SCE sweeping through +0.5 V vs Ag+/AgNO3. Two scans were run sequentially before the next addition of acid and the electrodes were cleaned in between each run. Following an addition of acid, the solution was bubbled with N2 for two minutes before performing another scan. As the catalytic wave increased and shifted to more negative potentials, the magnitude of the reducing potential was increased as necessary, up to -1.65 V vs Ag+/AgNO3. TOF’s were determined from the resulting voltammograms using the equation for kobs in the acid independent regime:  = 

1

#$%&'( 0.446 " )*

where icat is the current of the catalytic wave, ip is the peak current with one equivalent added acid, R is the gas constant, T is the temperature in Kelvin, F is Faraday’s constant and υ is the scan rate in mV/s.33,39,54 Plots of icat/ ip vs [4-cyanoanilinium] were made (Figure S3-8). Once icat/ ip reached a plateau, this value was used to calculate kobs (or the TOF in s-1). The kinetic isotope effect was observed in CV traces obtained using a 1 M HCl solution, made in either 10 mL of H2O or D2O to titrate a 10-3 M solution of catalyst in a solution of H2O or D2O, respectively. The experiment was performed using the same scan rate and applied potential as above (Figure S15). Determination of pKa An aqueous 1 M solution of NaOH was added to a 7.5 x 10-4 M solution of catalyst in 1:1 EtOH/H2O until the catalyst was completely deprotonated as monitored by UV-vis using a StellarNet SILVER-Nova25 BW16 Spectrometer. The sample was then titrated with 1 M or 0.1 M HCl, taking UV-Vis absorption spectra and recording pH between each addition, until the

ACS Paragon Plus Environment

25

The Journal of Physical Chemistry 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 26 of 35

catalyst was completely protonated. To address the error introduced by concentration differences in each measurement, the ratio of the catalyst’s absorbance at two different wavelengths was plotted against pH (Figure S3-8). The pKa of the catalyst was obtained by the inflection point of a fitted curve which was found by taking the second derivative of the equation of the curve. AUTHOR INFORMATION Corresponding Author *T.M. McCormick, [email protected]. Portland State University, Department of Chemistry, 1719 SW 10th Ave, Portland, Oregon, USA. 97201. Author Contributions The manuscript was written through contributions of all authors. All authors have given approval to the final version of the manuscript. Supporting information available Supporting information contains electrochemistry data, UV-vis absorbance spectra, structure energies, reaction coordinate diagrams, optimized structures, orbital maps and examples calculations. Funding Sources This work was funded by Portland State University. ACKNOWLEDGMENT Calculations were performed on the Research Computing Cluster, Hydra, at PSU. Mass spectroscopy was performed by Austin Shigemoto at Portland State University. REFERENCES

ACS Paragon Plus Environment

26

Page 27 of 35 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

The Journal of Physical Chemistry

(1)

Wiedner, E. S.; Appel, A. M.; Dubois, D. L.; Bullock, R. M. Thermochemical and Mechanistic Studies of Electrocatalytic Hydrogen Production by Cobalt Complexes Containing Pendant Amines. Inorg. Chem. 2013, 52, 14391–14403.

(2)

Helm, M. L.; Stewart, M. P.; Bullock, R. M.; Dubois, M. R.; Dubois, D. L. A Synthetic Nickel Electrocatalyst with a Turnover Frequency Above 100,000 s-1 for H2 Production. Science. 2011, 863–866.

(3)

Han, Z.; Eisenberg, R. Fuel from Water: The Photochemical Generation of Hydrogen from Water. Acc. Chem. Res. 2014, 47, 2537–2544.

(4)

McKone, J. R.; Marinescu, S. C.; Brunschwig, B. S.; Winkler, J. R.; Gray, H. B. EarthAbundant Hydrogen Evolution Electrocatalysts. Chem. Sci. 2014, 5, 865–878.

(5)

Du, P.; Eisenberg, R. Catalysts Made of Earth-Abundant Elements (Co, Ni, Fe) for Water Splitting: Recent Progress and Future Challenges. Energy Environ. Sci. 2012, 5, 6012.

(6)

Thoi, V. S.; Sun, Y.; Long, J. R.; Chang, C. J. Complexes of Earth-Abundant Metals for Catalytic Electrochemical Hydrogen Generation under Aqueous Conditions. Chem. Soc. Rev. 2013, 42, 2388–2400.

(7)

Rosser, T. E.; Gross, M. A.; Lai, Y.-H.; Reisner, E. Precious-Metal Free Photoelectrochemical Water Splitting with Immobilised Molecular Ni and Fe Redox Catalysts. Chem. Sci. 2016, 7, 4024–4035.

(8)

Chen, S.; Ho, M.; Bullock, R. M.; Daniel, L.; Dupuis, M.; Rousseau, R.; Raugei, S. Computing Free Energy Landscapes : Application to Ni-Based Electrocatalysts with Pendant Amines for H2 Production and Oxidation Computing Free Energy Landscapes :

ACS Paragon Plus Environment

27

The Journal of Physical Chemistry 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 28 of 35

Application to Ni-Based Electrocatalysts with Pendant Amines for H2 Production and Oxidation. ACS Catal. 2013, 4, 229–242. (9)

Solis, B. H.; Hammes-Schiffer, S. Theoretical Analysis of Mechanistic Pathways for Hydrogen Evolution Catalyzed by Cobaloximes. Inorg. Chem. 2011; Vol. 50, 1125211262.

(10)

Hammes-Schiffer, S. Catalysts by Design: The Power of Theory. Acc. Chem. Res. 2017, 50, 561–566.

(11)

Solis, B. H.; Yu, Y.; Hammes-Schiffer, S. Effects of Ligand Modification and Protonation on Metal Oxime Hydrogen Evolution Electrocatalysts. Inorg. Chem. 2013, 52, 6994– 6999.

(12)

Solis, B. H.; Hammes-Schiffer, S. Substituent Effects on Cobalt Diglyoxime Catalysts for Hydrogen Evolution. J. Am. Chem. Soc. 2011, 133, 19036–19039.

(13)

Raugei, S.; DuBois, D. L.; Rousseau, R.; Chen, S.; Ho, M.-H.; Bullock, R. M.; Dupuis, M. Toward Molecular Catalysts by Computer. Acc. Chem. Res. 2015, 48, 248–255.

(14)

Hammes-Schiffer, S. Proton-Coupled Electron Transfer: Moving Together and Charging Forward. J. Am. Chem. Soc. 2015, 137, 8860–8871.

(15)

Anson, C. W.; Ghosh, S.; Hammes-Schiffer, S.; Stahl, S. S. Co(salophen)-Catalyzed Aerobic Oxidation of P-Hydroquinone: Mechanism and Implications for Aerobic Oxidation Catalysis. J. Am. Chem. Soc. 2016, 138, 4186–4193.

(16)

Solis, B. H.; Hammes-Schiffer, S. Proton-Coupled Electron Transfer in Molecular Electrocatalysis : Theoretical Methods and Design Principles. Inorg. Chem. 2014, 53,

ACS Paragon Plus Environment

28

Page 29 of 35 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

The Journal of Physical Chemistry

6427–6443. (17)

Chen, S.; Rousseau, R.; Raugei, S.; Dupuis, M.; DuBois, D. L.; Bullock, R. M. Comprehensive Thermodynamics of Nickel Hydride Bis(diphosphine) Complexes: A Predictive Model through Computations. Organometallics 2011, 30, 6108–6118.

(18)

Panetier, J. A.; Letko, C. S.; Tilley, T. D.; Head-Gordon, M. Computational Characterization of Redox Non-Innocence in Cobalt-Bis(Diaryldithiolene)-Catalyzed Proton Reduction. J. Chem. Theory Comput. 2016, 12, 223–230.

(19)

De Proft, F.; Forquet, V.; Ourri, B.; Chermette, H.; Geerlings, P.; Morell, C. Investigation of Electron Density Changes at the Onset of a Chemical Reaction Using the State-Specific Dual Descriptor from Conceptual Density Functional Theory. Phys. Chem. Chem. Phys. 2015, 17, 9359–9368.

(20)

Liu, S.; Ess, D. H.; Schauer, C. K. Density Functional Reactivity Theory Characterizes Charge Separation Propensity in Proton-Coupled Electron Transfer Reactions. J. Phys. Chem. A 2011, 115, 4738–4742.

(21)

Tognetti, V.; Morell, C.; Joubert, L. Quantifying Electro/nucleophilicity by Partitioning the Dual Descriptor. J. Comput. Chem. 2015, 36, 649–659.

(22)

Han, Z.; Shen, L.; Brennessel, W. W.; Holland, P. L.; Eisenberg, R. Nickel Pyridinethiolate Complexes as Catalysts for the Light-Driven Production of Hydrogen from Aqueous Solutions in Noble-Metal-Free Systems. J. Am. Chem. Soc. 2013, 135, 14659–14669.

(23)

Han, Z.; Qiu, F.; Eisenberg, R.; Holland, P. L.; Krauss, T. D. Robust Photogeneration of

ACS Paragon Plus Environment

29

The Journal of Physical Chemistry 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 30 of 35

H2 in Water Using Semiconductor Nanocrystals and a Nickel Catalyst. Science 2012, 338, 1321–1324. (24)

Virca, C. N.; McCormick, T. DFT Analysis into the Intermediates of Nickel Pyridinethiolate Catalysed Proton Reduction. Dalt. Trans. 2015, 44, 14333–14340.

(25)

Frisch, M. J.; Trucks, G. W.; Schlegel, H. B.; Scuseria, G. E.; Robb, M. A.; Cheeseman, J. R.; Scalmani, G.; Barone, V.; Mennucci, B.; Petersson, G. A., et al. Gaussian 09, Revision D.01, Gaussian, Inc., Wallingford CT, 2009.

(26)

Becke, A. D. Density-Functional Exchange-Energy Approximation with Correct Asymptotic-Behavior. Phys. Rev. 1988, A, 3098–3100.

(27)

Miehlich, B.; Savin, A.; Stoll, H.; Preuss, H. Results Obtained with the CorrelationEnergy Density Functionals of Becke and Lee, Yang and Parr. Chem. Phys. Lett 1989, 157, 200–206.

(28)

Perdew, J. P. Density-Functional Approximation for the Correlation Energy of the Inhomogeneous Electron Gas. Phys. Rev. B 1986, 33, 8822–8824.

(29)

Barone, V.; Cossi, M. Quantum Calculation of Molecular Energies and Energy Gradients in Solution by a Conductor Solvent Model. J. Phys. Chem. A 1998, 102, 1995–2001.

(30)

Cossi, M.; Rega, N.; Scalmani, G.; Barone, V. Energies, Structures, and Electronic Properties of Molecules in Solution with the C-PCM Solvation Model. J. Comp. Chem. 2003, 24, 669–681.

(31)

Fourmond, V.; Jacques, P. A.; Fontecave, M.; Artero, V. H2evolution and Molecular Electrocatalysts: Determination of Overpotentials and Effect of Homoconjugation. Inorg.

ACS Paragon Plus Environment

30

Page 31 of 35 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

The Journal of Physical Chemistry

Chem. 2010, 49, 10338–10347. (32)

Felton, G. A. N.; Glass, R. S.; Lichtenberger, D. L.; Evans, D. H. Iron-Only Hydrogenase Mimics. Thermodynamic Aspects of the Use of Electrochemistry to Evaluate Catalytic Efficiency for Hydrogen Generation. Inorg. Chem. 2006, 45, 9181–9184.

(33)

Rountree, E. S.; McCarthy, B. D.; Eisenhart, T. T.; Dempsey, J. L. Evaluation of Homogeneous Electrocatalysts by Cyclic Voltammetry. Inorg. Chem. 2014, 53, 998310002.

(34)

Roberts, J. A. S.; Bullock, R. M. Direct Determination of Equilibrium Potentials for Hydrogen Oxidation/production by Open Circuit Potential Measurements in Acetonitrile. Inorg. Chem. 2013, 52, 3823–3835.

(35)

Savéant, J.-M. Molecular Catalysis of Electrochemical Reactions. Mechanistic Aspects. Chem. Rev. 2008, 108, 2348–2378.

(36)

Savéant, J. M.; Vianello, E. Potential-Sweep Voltammetry: General Theory of Chemical Polarization. Electrochim. Acta 1967, 12, 629–646.

(37)

Saveant, J. M.; Vianello, E. Potential-Sweep Chronoamperometry: Kinetic Currents for First-Order Chemical Reaction Parallel to Electron-Transfer Process (Catalytic Currents). Electrochim. Acta 1965, 10, 905–920.

(38)

Saveant, J. M.; Su, K. B. Homogenous Redox Catalysis of Electrochemical Reaction. J. Electroanal. Chem 1984, 171, 341–349.

(39)

Helm, M. L.; Stewart, M. P.; Bullock, M.; Dubois, M. R.; Dubois, D. L. A Synthetic Nickel Electrocatalyst. Science. 2011, 333, 863–866.

ACS Paragon Plus Environment

31

The Journal of Physical Chemistry 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

(40)

Page 32 of 35

Hu, X.; Brunschwig, B. S.; Peters, J. C. Electrocatalytic Hydrogen Evolution at Low Overpotentials by Cobalt Macrocyclic Glyoxime and Tetraimine Complexes. J. Am. Chem. Soc. 2007, 129, 8988–8998.

(41)

Bigi, J. P.; Hanna, T. E.; Harman, W. H.; Chang, A.; Chang, C. J. Electrocatalytic Reduction of Protons to Hydrogen by a Water-Compatible Cobalt Polypyridyl Platform. Chem. Commun. 2010, 46, 958–960.

(42)

Kilgore, U. J.; Roberts, J. A. S.; Pool, D. H.; Appel, A. M.; Stewart, M. P.; Dubois, M. R.; Dougherty, W. G.; Kassel, W. S.; Bullock, R. M.; Dubois, D. L. [Ni(PPh2NC6H4X2)2]2 Complexes as Electrocatalysts for H2 Production: Effect of Substituents, Acids, and Water on Catalytic Rates. J. Am. Chem. Soc. 2011, 133, 5861–5872.

(43)

Appel, A. M.; Pool, D. H.; O’Hagan, M.; Shaw, W. J.; Yang, J. Y.; Rakowski Dubois, M.; Dubois, D. L.; Bullock, R. M. [Ni(PPh2NBn2)2(CH3CN)]2+ as an Electrocatalyst for H2 Production: Dependence on Acid Strength and Isomer Distribution. ACS Catal. 2011, 1, 777–785.

(44)

Stubbert, B. D.; Peters, J. C.; Gray, H. B. Rapid Water Reduction to H2 Catalyzed by a Cobalt Bis(iminopyridine) Complex. J. Am. Chem. Soc. 2011, 133, 18070–18073.

(45)

McCrory, C. L.; Uyeda, C.; Peters, J. C. Electrocatalytic Hydrogen Evolution in Acidic Water with Molecular Cobalt Tetraazamacrocycles. J. Am. Chem. Soc. 2012. 134, 31643170.

(46)

O’Hagan, M.; Shaw, W. J.; Raugei, S.; Chen, S.; Yang, J. Y.; Kilgore, U. J.; Dubois, D. L.; Bullock, R. M. Moving Protons with Pendant Amines: Proton Mobility in a Nickel

ACS Paragon Plus Environment

32

Page 33 of 35 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

The Journal of Physical Chemistry

Catalyst for Oxidation of Hydrogen. J. Am. Chem. Soc. 2011, 133, 14301–14312. (47)

Rakowski DuBois, M.; DuBois, D. L. The Roles of the First and Second Coordination Spheres in the Design of Molecular Catalysts for H2 Production and Oxidation. Chem. Soc. Rev. 2009, 38, 62–72.

(48)

Yang, J. Y.; Bullock, R. M.; DuBois, M. R.; DuBois, D. L. Fast and Efficient Molecular Electrocatalysts for H2 Production: Using Hydrogenase Enzymes as Guides. MRS Bull. 2011, 36, 39–47.

(49)

Goff, A. L.; Artero, V.; Jousselme, B.; Tran, P.; Guillet, N.; Metaye, R.; Fihri, A.; Palacin, S.; Fontecave, M. From Hydrogenases to Noble Metal–Free Catalytic Nanomaterials for H2 Production and Uptake. Science. 2009, 326, 1384–1387.

(50)

Rountree, E. S.; Martin, D. J.; Mccarthy, B. D.; Dempsey, J. L. Linear Free Energy Relationships in the Hydrogen Evolution Reaction: Kinetic Analysis of a Cobaloxime Catalyst. ACS Cat. 2016, 6, 3326-3335.

(51)

Wiedner, E. S.; Brown, H. J. S.; Helm, M. L. Kinetic Analysis of Competitive Electrocatalytic Pathways: New Insights into Hydrogen Production with Nickel Electrocatalysts. J. Am. Chem. Soc. 2016, 138, 604-616.

(52)

Costentin, C.; Savøant, J. Multielectron , Multistep Molecular Catalysis of Electrochemical Reactions: Benchmarking of Homogeneous Catalysts. ChemElectroChem 2014, 1, 1226–1236.

(53)

Costentin, C.; Drouet, S.; Robert, M.; Save, J. Turnover Numbers, Turnover Frequencies, and Overpotential in Molecular Catalysis of Electrochemical Reactions. Cyclic

ACS Paragon Plus Environment

33

The Journal of Physical Chemistry 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 34 of 35

Voltammetry and Preparative-Scale Electrolysis. J. Am. Chem. Soc. 2012, 134, 11235– 11242. (54)

Wang, M.; Chen, L.; Sun, L. Recent Progress in Electrochemical Hydrogen Production with Earth-Abundant Metal Complexes as Catalysts. Energy Environ. Sci. 2012, 5, 6763.

(55)

Hansch, C.; Leo,

a; Taft, R. W. A Survey of Hammett Substituent Constants and

Resonance and Field Parameters. Chem. Rev. 1991, 91, 165–195.

ACS Paragon Plus Environment

34

Page 35 of 35 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

The Journal of Physical Chemistry

TOC Figure:

ACS Paragon Plus Environment

35